Very cool. Granted, I am a biased native Texan. I look forward to checking out these collections. UNT Digital Collections
The University of North Texas (UNT) has more than 1.5 million objects in its Digital Collections. It started collection and generating digital content in 2002, and it is part of the Portal to Texas History (300 partner institutions and private collectors), and the Gateway to Oklahoma History, in addition to its own UNT Digital Library.
It sounds like they basically designed their entire system in-house, using open source everything, and they are constantly upgrading services, features, and components as needed.
They use locally-qualified Dublin Core for their metadata, with 21 fields, regardless of collection, resource type, or source. The 22nd field is an administrative field.
All kinds of people enter metadata, from student workers to library professionals. They have written extensive documentation to insure quality and consistency. They use controlled vocabulary and field values, and name authority control.
They have provisions for share-ability and search-ability built in.
The Discussion section at the end of the article was quite good, with advice for others developing a system.
- Facets are pulled from controlled vocabularies or standardized formatting.
- Maintain a consistency with metadata schema and usage across holdings.
- Use authority control and controlled vocabularies. They can be based on established lists.
- Written documentation.